JVC Autocal Software V12 calibration for 2019 RS1000/RS2000/RS3000/NX5/NX7/NX9

[Manni]

Hi Manni, I take it your user1 calibration as above is for SDR viewing and your user2 for HDR ?

Any reason to not calibrate Natural or HDR10 modes instead ?
 

User 1 is for SDR rec-709
User 2 is for DCI-P3 for madVR HDR to SDR tonemapping

I use natural for HDTV (sdr 709 with CMD low) as I use it mostly for sports
I don't use HDR10, especially now that we have a DTM mode that's far superior (yet still not as good as madVR). That's the mode I use for streaming UHD HDR content from my UB900, until I get the Envy.

The only use for HDR10 is to run an autocal, as you can't autocal DTM directly, though it seems to me that it works just as well if not better to autocal my user2 (DCI-P3).

I'll report back when I find the time to run more tests re DTM calibration.

[CyberPaul]

Has anyone run autocal using the HDR10 Picture Mode and verified the results using independent software (HCFR, CalMAN etc)?

I restored my Init file to investigate an issue with Auto Iris, which brought back the blue gamma droop (manifesting as an increasing colour temperature as input level drops; see first attachment).  After running gamma autocal the After log shows the issue fixed (see second attachment), but the HCFR grey scale shows no improvement at all with the blue gamma droop.

I ended up re-running gamma autocal using the Natural and Cinema Picture Modes, and then the gamma droop got fixed for HDR10, with and without filter.

I had a similar issue with the e-shift models - some calibration results did not "stick".

I experienced something very similar using Spyder 5 for gamma and i1 pro2 with autocal 11 v1.20.
I performed an autocal with HDR10 and I got a very bad color temperature behavior like you described (please see first picture).
After having read your message, I tried to calibrate same combination (High lamp, Wide filter, Aperture = -1) but without using HDR10... And I got an almost perfect alignment (see second picture).
It looks like HDR10 has to be avoided during calibration. After all what really matters is the combination of parameters, as the manual says. I had a suspect that something was wrong when I saw that the post-cal log screens were showing as reference a 2.2 line for the gamma and the REC709 triangle for the color.

[riddle]

Hi Manni, you've got a download somewhere your color profile for (DCI-P3, 2.2, D65, low lamp, filter, iris open).  I am just now experimenting with MadVR and Lumagen Radiance PRO and i think that the DCI-P3 profile could be a better option for me if i will use SDR tonemapping too.

Thank you

[Manni]

Hi Manni, you've got a download somewhere your color profile for (DCI-P3, 2.2, D65, low lamp, filter, iris open).  I am just now experimenting with MadVR and Lumagen Radiance PRO and i think that the DCI-P3 profile could be a better option for me if i will use SDR tonemapping too.

Thank you

Hi Riddle,

You don't need a custom profile with these new models as long as you plan to use the filter, and it's much better to use the factory one. Simply select DCI instead of rec-709 or BT2020.

P3 is only preferable if you use a 3D LUT, to limit posterization. If you don't use a 3D LUT, it makes more sense to use a BT2020/2.4 baseline, because that's compatible with the tonemapping of consumer sources (such as a bluray player) if you want to compare madVR tonemapping with other sources, or have non-HTPC sources.

Either way, there is no need for custom profiles in the new models, as they have DCI-P3-F (DCI) and both BT-2020-F ( BT2020) and BT-2020-NF (HDR) options in the factory profiles. You would only need a custom profile if you wanted to use DCI-P3-NF (no filter) or Rec-709-F (filter), as these are not yet implemented in the firmware.

Happy New Year everyone!

[riddle]

Hi Manni, yes that's exactly what i want to test. I would like to compare playback via MadVR HTPC VS Lumagen Radiance PRO and decide what to use and for this reasons i will try DCI-P3-F (DCI).

Thnak you

[Manni]

Hi Manni, yes that's exactly what i want to test. I would like to compare playback via MadVR HTPC VS Lumagen Radiance PRO and decide what to use and for this reasons i will try DCI-P3-F (DCI).

Thnak you

You're welcome. Again, if you don't use a 3D LUT, BT2020 will give as good or better results as DCI, as long as you autocal it properly and the saturations are tracking well. You can measure DCI within BT2020 to check for this.

If your native gamut covers more than DCI-P3, you might get better results with BT2020.

If your native gamut doesn't cover more than DCI-P3, it shouldn't make any difference, unless you use a 3D LUT, in which case you'll most likely get better results with DCI-P3.

As long as saturations track properly and you don't use a 3D LUT over the Autocal, it's better to use the gamut that has the widest cover, as it might give you better color with the few titles that are mastered to BT2020 (there are quite a few now), assuming there is content beyond DCI-P3 in these titles, which is another question...

[riddle]

With my JVC N7 i can get somethink closer to 96-97% on DCI-P3. Manni which 3D LUT you are using on your MadVR?

Thank you

[Manni]

With my JVC N7 i can get somethink closer to 96-97% on DCI-P3. Manni which 3D LUT you are using on your MadVR?

Thank you

Then DCI-P3 is fine. I use a Lightning LUT made with Calman (about 100 points,  under 10 minutes). After an Autocal, there is no need for a bigger LUT, as long as you cover near to 100% of the target gamut, which is your case.

[estabanj]

Anyone had experience where importing the init file has not returned the projector to the original brightness?

This is on a NX7 running 3.1 software and V11 (I am still using a Spyder5Pro so I can't use V12)

I have tried powering off the PJ after the import and that doesn't make any difference. This is all with the projector having less than 60hours on the lamp.

When I first set the PJ up for SDR I would get 85 nits low lamp and -6 aperture for 60nits. Now even after importing the init file, I get 75 nits and well under 50 nits at -6. High lamp is similarly down from 127 nits to 114 nits.

Any ideas? I think I saw this with my previous NX7 too, but I wasn't as precise with my measurements last time, so I can't be sure.

Thanks

[Manni]

Restoring the init file doesn’t restore the initial brightness, it restores the factory calibration tables. Your lamp could have lost a significant amount of brightness over the first 60 hours. It could even get some of that brightness back in the next 60 hours. I have measured JVCs at a few hundred hours that had more brightness than when new. It’s pretty much unpredictable.

Losing 10% in 60 hours is more than usual if it’s only due to the bulb aging (and not to calibration), but it can happen, especially with long sessions in high lamp in a non temp controlled room. It should be fairly stable for a few hundred hours once it’s gone through its initial drop.

However, make sure that you also reset the RGB gains settings, as the init file only restores the calibration tables, it doesn’t touch the settings. So if you had set the gains to reach D65, you need to reset the gains setting after restoring the init file, or do a factory reset (which resets the settings but not the calibration tables).

[estabanj]

Hi Manni, thanks very much for the response.

Yes I use high lamp for HDR so that would fit within your experience. 
It is good to know this is somewhat normal.

Cheers Stephen 

[dcchan2]

No, it's not the complete state of the projector, as explained in my former post. It's only the complete calibration tables (all the calibration tables, so the current state of the projector as far as internal calibration tables are concerned, not just the last one). It doesn't include any of the settings.

One behaviour that puzzled me is that restoring one of my own back-up file takes a few seconds, whereas restoring the Init file takes about 30 minutes.  What additional “work” does autocal do with the Init file?

[SirMaster]

One behaviour that puzzled me is that restoring one of my own back-up file takes a few seconds, whereas restoring the Init file takes about 30 minutes.  What additional “work” does autocal do with the Init file?

INIT is all the calibration tables.

Your own back up file is just for 1 user mode as far as I know.

[dcchan2]

Here is a list of suggestions for future versions that would significantly improve the use of the software.
Feel free to contribute if you have any (reasonable) suggestion. I'll try to get this to JVCs through my contacts once we've gathered enough feedback.

The ability to save the Before and After logs would be useful.   Currently they just scroll by as the calibration progresses with no provisions to scroll back.

[Manni]

The ability to save the Before and After logs would be useful.  Currently they just scroll by as the calibration progresses with no provisions to scroll back.

Added to the list. I agree it would be nice, but I wouldn't hold my breath. There are too many more important improvements on the list. But you never know, sometimes, easy stuff get added before more important changes...